ashe

on the frontier of consumer wearables: building an egoless AI product in conversation with @minafahmi, building @sandbar between Mina's eloquence and our nyc galavanting, this conversation feels like a work of art. stoked to be able to share it dive in: 00:34 Mina's background 01:42 Robotics, neural interfaces, and hardware 06:06 Bringing nervous system work into designing AI products 10:25 Building with voice 12:38 AI extending self-concept 19:05 AI sycophancy 24:58 Agentic memory 30:11 Hardware decisions at Sandbar 34:16 Hiring at Sandbar 36:04 Bringing AI to more people

@DBredvick folks (esp nontechnical) are not using it as a prototype, they're using it earnestly as the standalone agent team and second brain sys. but there are loads of issues wrote a bit more about this after running a bunch of 1:1 lightning chats in april! x.com/ashebytes/stat…

@ashebytes including the claw

Rumored today that both OAI + Anthropic are in some way funding consulting firms that help companies adopt AI. This is neither bearish or bullish — just proves that it's going to be a 10-20 year slog to get all businesses agentified. Wrote about this a while back: drew.tech/posts/agi-some…

@DBredvick interesting, agreed mostly.. but when it gets down to the level of openclaw or the local second brain chat/obsidian set ups I've seen.. and the agent is given so much access, seems painful to sanitize down the line

@ashebytes most vibe coded apps are just good PRDs that should be rebuilt

@drewpusateri @OpenAI what are your thoughts now on threads? are you continuing to spend any time on there?

I’ve joined @OpenAI to work on the communications team. I've been off this platform for a couple years (RIP trying to make Threads happen), so excited to return to the calmest, healthiest possible online ecosystem.

@BusDownBonnor borrowing "take care"

Claude literally just ended the conversation on me???? This might be AGI

more info/cal here: ashe.ai/agency

going to practice on mondays being 10% more self promotional on here.. somewhat to my surprise, I am really enjoying coaching very ambitious founders/project leads on achieving large, creative leaps in tech if you’re working on a moonshot tech project, and looking for an injection of energy, alignment, network advice, & hands on tech/product conversations as you make large leaps to bring things to life.. I’ve been there and have become opinionated on what works and doesn’t. there’s always a way to win from a place of joy and abundance. there are more founders than ever and incentives to do your own thing, but nothing big happens without support - the key thing is to surround yourself with expanders that get the vision, lift you up, and make you better it’s been personally very energizing for me to be that teammate. shoot me a dm

agreed to actually push the needle and deliver value imo requires 1. experience on a founding team with bringing agentic products to market and *actually* being in the reality of working with users and delivering value specific to their workflows ^ not easy. a lot of folks don’t engage in the actual feedback loops of “is this being used? Is this being used as expected? what is the treatment effect? what is the path to scaling up value 2x,10x etc? 2. coordination and comms within the client team also very hard! you must be able to speak their language, understand the teams incentive structure (not always towards optimal/efficient outcomes), and actually integrate the solution and its distribution or have a strong handoff that makes this possible

@ashebytes other question is where you find enough people qualified to advise on and build these solutions properly… how do you staff the firms?

can’t entirely tell but it seems OpenAI and anthropic are each building their own AI native Accenture? curious who they’re going to staff these firms with… and whether they’re allowed to use models or offerings from other companies in their solutions?

@dejavucoder love this

one more since you guys are enjoying "Lets draw a poster of the policy gradient equation to intuitively explain it with minimal text and relevant notation. the text should be readable. Fusion of Bauhaus and Risograph." bauhaus gives the shapes.

@Aabbhhz know 👏which 👏game👏you’re👏in

@ashebytes incredible game this.

everyone's looking for that sama follow,, but my friends have other plans for me

@TheAnnaGat will be keeping an eye out for these fits

@TheAnnaGat loove it

@alexkehr ?

people weren’t lying about current SF rent being out of control… the only place for rent in my neighborhood right now is 24k/month

@erikalee dms are open 👀

@erikalee they need AI too, I wanna make em AI native

@sheep_xxl @sama nah feels like he’s just shooting em off

@ashebytes @sama Do you think he uses ai to style his tweets?

you know I am digging the more playful @sama posting energy honestly all of us in AI could use more silly banter..ultimately we’re all in this together might as well have a lot of fun with it